ggTrader
An algorithmic trading bot and a research framework for professional strategy development.
Mission Objective
Create a reproducible and scalable trading research and execution framework for professional traders.
Project Overview
ggTrader is a professional algorithmic trading framework designed for high-performance research and execution. It moves away from monolithic scripts toward a modular, scalable architecture that supports complex strategy validation and multi-exchange connectivity.
Core Features
- Modular Architecture: Clean separation between core engine logic, portfolio management, signal indicators, and exchange adapters (Kraken).
- Reproducible Research: A dedicated
ResultsManagerand timestamped results folders ensure that every backtest and optimization run is tracked and auditable. - Walk-Forward Optimization (WFO): Finds stable parameters over sliding time windows to reduce overfitting.
- Sensitivity Analysis: Tests how strategy performance reacts to parameter drift, ensuring robustness in changing markets.
- Professional Analytics: Deep integration with Jupyter Notebooks for interactive visualization and performance deep-dives.
Technical Architecture
- Core: Python, NumPy, Pandas
- Backtesting: High-performance simulation logic (integrated with VectorBT patterns)
- Data: Parquet local storage, Kraken API adapters
- Optimization: Custom WFO and sensitivity analysis pipelines
Project Log
No log entries found for this project.